Reports

Report Number: 131
Year: 2012
 

Continuing Calibration and Application of LUOM in the Southern Guam Watersheds.

In Southern Guam, there are few watersheds with both rainfall and streamflow data. Other watersheds have only rainfall data but no streamflow data. For those watersheds without streamflow data, it is obviously difficult to carry out watershed management studies which require streamflow data. Meanwhile, there are two problems with most of the watersheds with streamflow data. One is that the streamflow gage is not always located at the watershed outlet but at a distance upstream of the outlet. The other is that there are many missing data in the streamflow record. These problems create difficulties in the development of optimal watershed management plans.

The Large-scale, Unified and Optimization Model, LUOM (Luo, 2007) is a fully physically based, two-dimensional distributed watershed model which simulates the hydrologic cycle on a watershed scale. The model discretizes the watershed into rectangular grid cells and makes use of spatial distributed GIS (Geographic Information Systems) data such as DEM (Digital Elevation Model), vegetation, and soil data. The model is comprised of a series of sub-models for climate data distribution, evapotranspiration, infiltration, groundwater flow, surface flow, etc. The surface flow sub-model solves the two-dimensional shallow water equations using the diffusive wave approximation. With the input of climatic data, mainly precipitation, temperature and wind speed, the model is able to generate not only one-dimensional output – discharge hydrographs, but also two-dimensional hydrologic quantities such as evapotranspiration, infiltration, soil moisture, groundwater table and surface water depth. Simulation of the impacts of land use (vegetation) transformation and global climate changes is within the model’s capability.

The objective of this study was to continue to: 1) Calibrate and validate the watershed model – LUOM (Luo, 2007) in the Southern Guam watersheds in which there are both rainfall and streamflow data, and 2) Apply the calibrated models to the Southern Guam watersheds both with and without streamflow gages to generate long term time series of streamflow for the whole watershed. The watersheds for this project were those that were not covered by the preceding project in 2009 (Luo and Khosrowpanah, 2010).

In this study, the LUOM (Luo, 2007) has been calibrated and validated for the Talofofo, Ylig, Pago, Atantano and Finile watersheds. In the Talofofo watershed, there are eight USGS (United States Geological Survey) streamflow gages. The period of record for only four of these flow gages coincide with those of the rainfall data collected at the USGS/NCDC rain gages located either inside or close by the watershed. Therefore, the LUOM was calibrated and validated at these four streamflow gages in the Talofofo watershed. The calibrated models were applied to a total of 15 watersheds including the five calibration watersheds and the other 10 adjacent watersheds. The preceding project, that was completed in 2009, used the fifty four (54) years of rainfall data as input, to the model in order to generate long term time series of streamflow. The final output of the long term time series of streamflow studies is a combination of the observed streamflow data collected at the USGS gage, if the watershed had any, and the long term simulation result from the model.

Author(s):
Qiang Charles Luo
Shahram Khosrowpanah